Brett S. Duersch and Dr. Paul B. Farnsworth, Chemistry and Biochemistry

Inductively coupled plasma mass spectrometry (ICP-MS) is one of the fastest growing fields in analytical chemistry. This discipline combines the powerful detection of mass spectrometry (MS) with the almost universal ionization capabilities of the inductively coupled plasma (ICP) . Combing these two techniques is not a trivial matter, since ions must be extracted from a plasma operating at atmospheric pressures and at temperatures ranging from 6,000 to 10,000 C into a mass spectrometer which operates at vacuum pressures(10 -10 torr). At present the ICP and -3 -6 the MS are typically interfaced with a series of two cones. The first is placed in the plasma itself and the second a short distance behind it. This arrangement, combined with differential vacuum pumping, creates a beam of charged particles and ions traveling into the mass spectrometer at supersonic speeds, which is subsequently analyzed.

A great deal of work has been done to investigate the formation and subsequent behavior of the ion beam as it passes from the plasma, at atmospheric pressures into a vacuum chamber. These studies have had as their purpose the development of an improved instrument through an improved ion extraction design. As Chen and Farnsworth1 recently noted, there are several difficulties with experiments conducted thus far. Two significant difficulties have been that techniques are invasive, for example placing a charged probe 2 or graphite targets3 directly into the ion beam. A second significant problem is that the techniques are not species selective, in that they cannot distinguish between ions, charged particles nor different elements2,3. Because of the problems associated with these measurement methods, the results of these experiments are questionable1.

In our lab we are developing a novel technique for exploring the ion beam, that is noninvasive, species selective and can map a major portion of the ion beam. This method utilizes laser induced fluorescence from probes mounted inside of the mass spectrometer to map the ion beam. Laser light is carried into the mass spectrometer through a fiber optic which is then used to excite species in the ion beam. Resulting fluorescence photons are removed from the vacuum chamber for detection through another optical fiber. The fibers and necessary optics to focus laser light and collect fluorescence signals, hereafter referred to as the probes, are mounted to computer controlled positioning stages. This allows the probes to be moved in a coordinated effort to generate three dimensional maps of the ion beam.

For our initial experiments we selected Ba+ as the analyte. Ba+ has a strong nonresonant fluorescence scheme, thus the laser wavelength necessary to excite the Ba+ is different from the wavelength of fluorescence. This is an important measure to allow reduced scattered light and to obtain a good signal to background ratio. Furthermore Ba+ is inexpensive and poses no significant health hazards under normal laboratory handling procedures. Initial trials of the experimental setup were done by mapping the ion beam along its propagational axis, or the z-axis of the experiment. The results shown for two concentrations of Be are shown in Fig. 1. The y axis of the graph shows the number of fluorescence photons collected for every 3000 laser shots, the x axis shows the distance along the z axis from the tip of the second of the two cones described above, known as the skimmer cone. “Corrected” signals are the initial data collected for each concentration minus the blank signal, a solution of 1% HN03 containing no Ba . The curved portion at + the beginning of the curve does not indicate a smaller number of ions present but is caused by a lack of linear response from the detector at high signal counts.

There are several significant results from this data. The first is that our technique for ion beam mapping is valid. Counts should drop as the probes are moved farther from the skimmer as a result of reduced ion density from an expanding beam. The maps are concentration dependent such that the four times more concentrated solution has about four times as many counts as the less concentrated solution, just as we would expect. Most interesting is that our results here suggest that the ion density decays more rapidly than expected from theory4 and from other experiments2. At the present time further experiments with the laser probes are underway in our laboratory to better characterize and understand this deviation5.

References

1. Chen, Y. and P.B. Farnsworth, awaiting publication in Spectrochimica Acta.
2. Hongsen, N. and R. S. Houk, Spectrochimica Acta, 49B, 1283 (1994).
3. Chen, X. and R. S. Houk, Spectrochimica Acta, 51 B, 41 (1996).
4. Douglas, D.J., Inductively Coupled Plasmas in Analytical Atomic Spectrometry 2nd ed. Montaser and Golightly: VCH, New York, 630 (1992).
5. This research was supported under a grant from the National Science Foundation number CHE-9415384. The invaluable assistance of Yibai Chen and Dr. Adeline Ciocan is gratefully acknowledged.